THIS ANALYSIS INVOLVED THE CONSTRUCTION OF A COMMON MEASURE OF THE LEVEL OF DEMAND FOR COURT SERVICES AND OF THE LEVEL OF OUTPUT OF COURT SERVICES UTILIZING A MODIFIED VERSION OF THE CASE WEIGHTS DERIVED BY THE FEDERAL JUDICIAL CENTER. THESE WEIGHTS REFLECT THE NUMBER OF JUDGE MINUTES REQUIRED TO DISPOSE OF THE AVERAGE CASE IN EACH OF THE APPROXIMATELY 42 CLASSES OF CASES IN WHICH CASE DATA IS PUBLISHED AT THE DISTRICT COURT LEVEL. ANNUAL TOTALS OF CASE-RELATED JUDGE HOURS WERE CONVERTED INTO 'EQUIVALENT JUDGE YEARS' BY ASSUMING THAT A FULL-TIME JUDGE CAN SUPPLY 1302 CASE-RELATED HOURS PER TWELVE MONTHS OF SERVICE. THIS STUDY ALSO INVOLVED THE APPLICATION OF THE TECHNIQUES OF MULTIVARIATE STATISTICAL INFERENCE TO THE MEASUREMENT OF BEHAVIORAL RELATIONSHIPS AMONG ALL FEDERAL DISTRICT COURTS. STUDY FINDINGS REVEALED THAT JUDICIAL PRODUCTIVITY IN EQUIVALENT JUDGE YEARS OF OUTPUT PER AVAILABLE JUDGE HAS, ON THE AVERAGE, INCREASED BY 32 PERCENT BETWEEN 1968 AND 1974; THAT THIS INCREASE IN AVERAGE JUDICIAL OUTPUT HAS RISEN FASTER THAN THE INCREASE IN AVAILABLE WORKLOAD; AND THAT THE SIMULTANEOUS EFFECTS OF PRODUCTIVITY AND DEMAND DIFFERENCES IN DISTRICT COURTS EXPLAIN ONLY ABOUT HALF OF THE VARIATION IN CASE PROCESSING TIMES AMONG DISTRICTS. THESE DIFFERENCES WERE EXPLAINED PRIMARILY BY DEMAND PER AVAILABLE JUDGE DERIVED FROM THE PENDING CASE LOAD AT THE START OF EACH YEAR. COUNTER TO EXPECTATIONS, COURTS WHICH RELIED MORE HEAVILY ON TRIALS AS DISPOSITION METHODS DID NOT SHOW LOWER LEVELS OF PRODUCTIVITY.